DELINEATION OF TOPOGRAPHIC AND HYDROGEOLOGIC CONTROLS ON LAKE SALINTY IN THE NEBRASKA SAND HILLS

The Sand Hills are the largest sand dune field (about 50,000 km2) in the western hemisphere. Numerous shallow lakes (mostly not deeper than 1 m) occur in topographic depressions under west-east regional shallow groundwater flow. In Sheridan and Garden counties alone there are approximately 400 lakes with surface areas larger than 4 ha. TDS in lake water ranges from 0.3 g/L to more than 100 g/L, and pH ranges from 8.4 to more than 10. The groundwater concentrations are typically low. Although several hypotheses are available, causes of wide variations in lake salinity within this large geographic area have not been determined conclusively. The objective of this study is to delineate mechanisms of salinization of these lakes using the concept of mass balance.

In conditions without surface water runoff in lake basins and direct groundwater-lake water connection, we based our approach on the following relationship between concentration of groundwater (Cg) and concentration of lake water (Cl): Cl=Cg (Qin/Qout), where Qin is groundwater inflow and Qout is groundwater outflow. The evaluation of crucial ratio of Qin/Qout can be based on various hydrodynamic models however these models are subject to broadly different underlying assumptions. One of the analytical models determines this ratio by using dimensionless parameter aUo/(kH) where a is lake radius, Uo is the regional flow Darcy's velocity, k is hydraulic conductivity and H is vertical shift of lake surface from regional water table. This model shows effects of lake geometry, regional groundwater flow, aquifer properties, topography and climate.

Validity of this model was assessed in Crescent Lake National Wildlife Refuge Area, Nebraska. Using data on aquifer properties from direct push techniques, well data for determining groundwater flow direction and magnitude, GIS for lake dimensions and lake position with respect to water table, we attempt to identify assumptions of this model and major parameters involved. Preliminary results show that this dimensionless parameter is correlated to lake salinity. Considering proportional relationship between salinity and Qin/Qout ratio, this model can be further explored using numerical modeling.